not a problematic file at all - fixed positions for all columns so its easy to process.
Do NOT use the import wizard - you need to manually process the file (SSIS can be hard, easier in T-SQL)
So define the file as a single column with the max size of each row and load onto a staging table.
from the staging table you select all rows that are
- Not empty
- are not like '======='
- do not have the particular header strings on the exact positions they are on
the remaining rows you can use substring and cast to convert to the individual columns as desired and insert into your final main table (or do further process with the data)